Journal of Chromatography, Elsevier Science Publishers
CHROM.
539 (1991) 177-185 B.V., Amsterdam
22 953
Separation of antibody-antigen complexes by capillary zone electrophoresis, isoelectric focusing and high-performance size-exclusion chromatography R. G. NIELSEN,
E. C. RICKARD,
Lilly Research Laboratories, (First received
P. F. SANTA,
Eli Lilly & Company,
June 6th, 1990; revised manuscript
D. A. SHARKNAS Indianapolis, received
and G. S. SITTAMPALAM*
IN 46285 (U.S.A.)
October
2nd, 1990)
ABSTRACT The separation of antibody-antigen complexes by free-solution capillary zone electrophoresis (CZE) has been demonstrated. The antigen, a monoclonal antibody specific for the antigen, and the complex were well resolved. The entire separation was achieved in less than 10 min using on-column UV detection. The pl values for the three species were estimated separately by isoelectric focusing (IEF) experiments on polyacrylamide gels. Reasonably good agreement was found between the relative migration times measured by CZE and the p1 values. Both IEF and high-performance size-exclusion chromatography of the antibody-antigen mixtures confirmed the formation of the complexes observed by CZE. This study demonstrates the utility of CZE as a new and complementary technique for the characterization of antibodyantigen complexes.
INTRODUCTION
Capillary zone electrophoresis (CZE) in free solution has gained widespread recognition as a powerful separation tool that is complementary to gel-electrophoretic and chromatographic techniques. Recent literature describes several applications, along with information on basic principles, instrumentation and methodology [l-l I]. This paper describes the separation of antibody-antigen complexes formed by an immunochemical reaction by free-solution CZE; the antigen is a recombinant human growth hormone (hGH) that reacts with a monoclonal antibody [MoAb, immunoglobulin G (IgG) class] specific to hGH. The complexes were well resolved from the unreacted (or free) antibody and antigen in solution. Recently, Chen et al. [12] also reported a successful separation of antibody-antigen complexes by CZE. However, they employed an isoelectric focusing experiment that requires a stabilizing medium and ampholytes in capillary tubes. In our work, the separation was achieved in a free-flowing buffer solution without ampholytes or stabilizing media. Preliminary results of this work were reported elsewhere [6]. We also present data comparing CZE to conventional isoelectric focusing (IEF) and high-performance size-exclusion chromatography (HPSEC). The monoclonal preparation employed in this study, at least in theory, contains 0021-9673/91/%03.50
0
1991 Elsevier
Science Publishers
B.V.
R. G. NIELSEN et al.
178
a single, homogeneous population of antibody molecules specific to a well-defined epitope on the antigen. Thus, only two types of complexes [IgG-hGH and IgG-(hGH)z] will be formed, that correspond to reaction at the two antigen binding sites of each antibody molecule: IgG + hGH Y IgG-hGH
IgG-hGH
+ hGH i=-’ IgG-(hGH)2
(I) (2)
We will refer to these complexes in general as IgG-(hGH),, where n is the valency of the antibody. These complexes are non-covalent in nature, tightly held together by various combinations of electrostatic, hydrophobic, dipole-dipole, hydrogen bonding and Van der Waals interactions [ 131. EXPERIMENTAL
Chemicals and reagents Highly purified recombinant DNA-derived hGH was obtained from Lilly Research Labs. (Indianapolis, IN, U.S.A.). The anti-hGH monoclonal antibody (part No. 20065) was purchased from Hybritech (San Diego, CA, U.S.A.). This preparation was a murine IgG made by ammonium sulphate fractionation and subsequent purification by DEAE ion-exchange chromatography. Calibration standards (~13-9) for IEF and the precast polyacrylamide gels were purchased from Pharmacia LKB Biotechnology (Piscataway, NJ, U.S.A.). Tricine was purchased from Sigma (St. Louis, MO, U.S.A.). All other chemicals were analytical-reagent grade, and the solutions were made in distilled water, deionized using the Mill&Q water purification system (Millipore, Bedford, MA, U.S.A.). Apparatus CZE experiments were performed using an Applied Biosystems (Foster City, CA, U.S.A.) Model 270A instrument. The fused-silica capillary was 100 cm long (80 cm to the detector) with 50 pm I.D. and 360 pm O.D. (Polymicro Technologies, Phoenix, AZ, U.S.A.). The capillary was rinsed with the 0.10 A4 tricine buffer (pH 8) between injections. New capillaries were rinsed with 0.10 M NaOH, water and 0.10 M HCI prior to use. In all experiments, the detector end of the capillary was negatively polarized relative to the injector end. The IEF experiments were carried out on the Pharmacia PhastSystem (Pharmacia LKB Biotechnology). HPSEC was performed using a high-performance liquid chromatographic system consisting of a HewlettPackard Model 1050 series gradient pump (Waldbron Analytical Division, F.R.G.), ISS-100 sampling system (Perkin-Elmer Norwalk, CT, U.S.A.) and a Spectroflow Model 757 absorbance detector (Kratos, Ramsey, NJ, U.S.A.). Electrophoretic conditions CZE separations were performed in 0.1 M tricine buffer, pH 8, using an applied voltage of 30 kV, with a current of about 19 PA. Injection volumes were ca. 9 nl in all cases (3-s injection using a 127-mmHg vacuum), and the separated proteins were detected at 200 nm. Operating temperature was maintained at 30.0 + O.l”C.
CZE, IEF AND HPSEC OF ANTIBODY-AWTIGEN
COMPLEXES
179
In the IEF experiments, I-PI volumes of solutions containing individual proteins or protein mixtures (prepared as described below) were directly applied to the precast polyacrylamide gels (5%) using special sample application combs provided for the Pharmacia PhastSystem. The gels were pre-focused for 75 V . h at 15”C, 2000 V, prior to sample application for 15 V h (200 V, 2.5 mA, 3.5 W). Focusing was carried out at 15°C for 410 V h during which time the voltage rose from 200 to 2000 V (2.5 mA, 3.5 W). Gels were fixed in 20% trichloroacetic acid and stained with Coomassie Blue R-350, following standard procedures described in the PhastSystem manual. Chromatographic
conditions
Separations were achieved with a GF-250 (25 cm x 9.4 mm) size-exclusion column (DuPont, Wilmington, DE, U.S.A.) at ambient temperature using a 0.20 M Na2HP04 (pH 7.6)mobile phase flowing at 0.60 ml/min. Eluted peaks were detected at 214 nm. Sample injection was carried out using a 20-~1 loop. Sample preparation
Both hGH and the anti-hGH MoAb (IgG) were dissolved directly into or dialyzed against 0.1 M tricine buffer (pH 8). The final concentration of IgG was 3.03 mg/ml(2.35. 10e5 M), and that of hGH was 0.508 mg/ml(2.29 lop5 M), based on molecular weights of 150 000 dalton for IgG and 22 124 dalton for hGH. These concentrations were determined by UV absorbance at 276 nm for hGH (a = 0.74 ml/mg cm) and IgG (a = 1.40 ml/mg . cm), where a is the absorbance of a 1 mg/ml solution in a 1 cm path length cell. Appropriate volumes of hGH and IgG solutions were mixed to initiate the reaction. The approximate molar ratios of 1gG:hGH in the mixtures used for CZE and HPSEC were 1:4, 1:2, I:], 2: 1 and 3:l. For the IEF experiments, hGH and the monoclonal IgG were dissolved in 0.05 A4 potassium phosphate buffer, pH 7.6. The final concentration of IgG was 2.2 mg/ml, and that of hGH was 1.Omg/ml. Two solutions containing IgG and hGH were made by mixing appropriate volumes of the above protein solutions. The approximate molar ratios of 1gG:hGH in the above mixtures were calculated to be 1:1.5 and 1:3.1. RESULTS
The model antigen (hGH) and the monoclonal anti-hGH antibody (IgG) were readily available in highly purified preparations. They were well characterized in-house to establish purity, identity and binding properties. Results [14] showed that both preparations contained a single protein component of appropriate molecular mass by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) (hGH cu. 22 124 dalton, IgG cu. 150 000 dalton). No significant protein impurities were detected when stained with Coomassie Blue R-350. The apparent affinity constant for the reaction (&) was cu. 1.0 . IO9 l/mol by cold competition radioimmunoassay [ 143.This strong binding affinity of the antibody for hGH results in strong, stable antibody-antigen complexes that form very rapidly in solution. Mixtures containing various molar ratios of 1gG:hGH were analyzed by CZE, IEF and HPSEC to study the properties of the complexes formed. These techniques separate species based on differences in electrophoretic mobility, isoelectric point and size, respectively.
R. G. NIELSEN et al.
180
0
IOO 200
300 400
500 600
700 600
900
Time(s) B
I
Free IGG in mixture), ‘1,
II Ii
IGG:HGH=3:1 IGG:HGH
= 2:l
IGG:HGH
= 19
IQGG
0
150
300
450
El
600
750
900
Time (s)
Fig. I. Separation of IgG, hGH and IgG-(hGH), complexes by CZE. (A) Electropherograms of IgG, hGH and mixtures containing an excess of hGH. (B) Electropherograms of IgG, hGH and mixtures containing an excess of IgG. Experimental conditions are described in the text.
CZE experiments Fig. 1 shows seven electropherograms (El-E7) representing separate injections. Electropherograms El and E2 are for individual injections of IgG and hGH, respectively. They show that these species have different migration times and are well resolved under the experimental conditions employed. Note that TgG migrates much faster than hGH, due to the fact that the plfor hGH is 5.2 and that for the monoclonal IgG is in the range 7.2-8.0 (multiple bands in Fig. 2). Therefore, hGH molecules have much greater negative charge than IgG molecules in the CZE separation buffer (0.1 M tricine, pH 8). This results in faster migration of the IgG molecules towards the negatively polarized detector end of the capillary. The first peak in all electropherograms is due to a slightly higher tricine concentration in the sample buffer compared to that of the separation buffer. Electropherograms E3 and E4 show the results for 1:4 and 1:2 molar ratios of IgG:hGH, respectively (Fig. 1A). Based on migration-time comparison, the peak at CL~. 531 s is a new molecular entity, most probably corresponding to one or more IgG+hGH), complexes. The peak at ea. 576 s corresponds to free hGH present in excess in both mixtures. As the 1gG:hGH molar ratios decrease in the mixtures, one observes distinct changes in the peak profiles (Fig. lB, electropherograms E5, E6 and E7). In electropherogram E5, the free IgG peak was observed at CLI.495 s along with two distinct peaks for IgG-(hGH), complexes at cu. 513 and ea. 531 s, respectively.
CZE, IEF AND HPSEC OF ANTIBODY-ANTIGEN
COMPLEXES
181
Fig. 2. Separation of IgG, hGH and IgG
CHROM.
539 (1991) 177-185 B.V., Amsterdam
22 953
Separation of antibody-antigen complexes by capillary zone electrophoresis, isoelectric focusing and high-performance size-exclusion chromatography R. G. NIELSEN,
E. C. RICKARD,
Lilly Research Laboratories, (First received
P. F. SANTA,
Eli Lilly & Company,
June 6th, 1990; revised manuscript
D. A. SHARKNAS Indianapolis, received
and G. S. SITTAMPALAM*
IN 46285 (U.S.A.)
October
2nd, 1990)
ABSTRACT The separation of antibody-antigen complexes by free-solution capillary zone electrophoresis (CZE) has been demonstrated. The antigen, a monoclonal antibody specific for the antigen, and the complex were well resolved. The entire separation was achieved in less than 10 min using on-column UV detection. The pl values for the three species were estimated separately by isoelectric focusing (IEF) experiments on polyacrylamide gels. Reasonably good agreement was found between the relative migration times measured by CZE and the p1 values. Both IEF and high-performance size-exclusion chromatography of the antibody-antigen mixtures confirmed the formation of the complexes observed by CZE. This study demonstrates the utility of CZE as a new and complementary technique for the characterization of antibodyantigen complexes.
INTRODUCTION
Capillary zone electrophoresis (CZE) in free solution has gained widespread recognition as a powerful separation tool that is complementary to gel-electrophoretic and chromatographic techniques. Recent literature describes several applications, along with information on basic principles, instrumentation and methodology [l-l I]. This paper describes the separation of antibody-antigen complexes formed by an immunochemical reaction by free-solution CZE; the antigen is a recombinant human growth hormone (hGH) that reacts with a monoclonal antibody [MoAb, immunoglobulin G (IgG) class] specific to hGH. The complexes were well resolved from the unreacted (or free) antibody and antigen in solution. Recently, Chen et al. [12] also reported a successful separation of antibody-antigen complexes by CZE. However, they employed an isoelectric focusing experiment that requires a stabilizing medium and ampholytes in capillary tubes. In our work, the separation was achieved in a free-flowing buffer solution without ampholytes or stabilizing media. Preliminary results of this work were reported elsewhere [6]. We also present data comparing CZE to conventional isoelectric focusing (IEF) and high-performance size-exclusion chromatography (HPSEC). The monoclonal preparation employed in this study, at least in theory, contains 0021-9673/91/%03.50
0
1991 Elsevier
Science Publishers
B.V.
R. G. NIELSEN et al.
178
a single, homogeneous population of antibody molecules specific to a well-defined epitope on the antigen. Thus, only two types of complexes [IgG-hGH and IgG-(hGH)z] will be formed, that correspond to reaction at the two antigen binding sites of each antibody molecule: IgG + hGH Y IgG-hGH
IgG-hGH
+ hGH i=-’ IgG-(hGH)2
(I) (2)
We will refer to these complexes in general as IgG-(hGH),, where n is the valency of the antibody. These complexes are non-covalent in nature, tightly held together by various combinations of electrostatic, hydrophobic, dipole-dipole, hydrogen bonding and Van der Waals interactions [ 131. EXPERIMENTAL
Chemicals and reagents Highly purified recombinant DNA-derived hGH was obtained from Lilly Research Labs. (Indianapolis, IN, U.S.A.). The anti-hGH monoclonal antibody (part No. 20065) was purchased from Hybritech (San Diego, CA, U.S.A.). This preparation was a murine IgG made by ammonium sulphate fractionation and subsequent purification by DEAE ion-exchange chromatography. Calibration standards (~13-9) for IEF and the precast polyacrylamide gels were purchased from Pharmacia LKB Biotechnology (Piscataway, NJ, U.S.A.). Tricine was purchased from Sigma (St. Louis, MO, U.S.A.). All other chemicals were analytical-reagent grade, and the solutions were made in distilled water, deionized using the Mill&Q water purification system (Millipore, Bedford, MA, U.S.A.). Apparatus CZE experiments were performed using an Applied Biosystems (Foster City, CA, U.S.A.) Model 270A instrument. The fused-silica capillary was 100 cm long (80 cm to the detector) with 50 pm I.D. and 360 pm O.D. (Polymicro Technologies, Phoenix, AZ, U.S.A.). The capillary was rinsed with the 0.10 A4 tricine buffer (pH 8) between injections. New capillaries were rinsed with 0.10 M NaOH, water and 0.10 M HCI prior to use. In all experiments, the detector end of the capillary was negatively polarized relative to the injector end. The IEF experiments were carried out on the Pharmacia PhastSystem (Pharmacia LKB Biotechnology). HPSEC was performed using a high-performance liquid chromatographic system consisting of a HewlettPackard Model 1050 series gradient pump (Waldbron Analytical Division, F.R.G.), ISS-100 sampling system (Perkin-Elmer Norwalk, CT, U.S.A.) and a Spectroflow Model 757 absorbance detector (Kratos, Ramsey, NJ, U.S.A.). Electrophoretic conditions CZE separations were performed in 0.1 M tricine buffer, pH 8, using an applied voltage of 30 kV, with a current of about 19 PA. Injection volumes were ca. 9 nl in all cases (3-s injection using a 127-mmHg vacuum), and the separated proteins were detected at 200 nm. Operating temperature was maintained at 30.0 + O.l”C.
CZE, IEF AND HPSEC OF ANTIBODY-AWTIGEN
COMPLEXES
179
In the IEF experiments, I-PI volumes of solutions containing individual proteins or protein mixtures (prepared as described below) were directly applied to the precast polyacrylamide gels (5%) using special sample application combs provided for the Pharmacia PhastSystem. The gels were pre-focused for 75 V . h at 15”C, 2000 V, prior to sample application for 15 V h (200 V, 2.5 mA, 3.5 W). Focusing was carried out at 15°C for 410 V h during which time the voltage rose from 200 to 2000 V (2.5 mA, 3.5 W). Gels were fixed in 20% trichloroacetic acid and stained with Coomassie Blue R-350, following standard procedures described in the PhastSystem manual. Chromatographic
conditions
Separations were achieved with a GF-250 (25 cm x 9.4 mm) size-exclusion column (DuPont, Wilmington, DE, U.S.A.) at ambient temperature using a 0.20 M Na2HP04 (pH 7.6)mobile phase flowing at 0.60 ml/min. Eluted peaks were detected at 214 nm. Sample injection was carried out using a 20-~1 loop. Sample preparation
Both hGH and the anti-hGH MoAb (IgG) were dissolved directly into or dialyzed against 0.1 M tricine buffer (pH 8). The final concentration of IgG was 3.03 mg/ml(2.35. 10e5 M), and that of hGH was 0.508 mg/ml(2.29 lop5 M), based on molecular weights of 150 000 dalton for IgG and 22 124 dalton for hGH. These concentrations were determined by UV absorbance at 276 nm for hGH (a = 0.74 ml/mg cm) and IgG (a = 1.40 ml/mg . cm), where a is the absorbance of a 1 mg/ml solution in a 1 cm path length cell. Appropriate volumes of hGH and IgG solutions were mixed to initiate the reaction. The approximate molar ratios of 1gG:hGH in the mixtures used for CZE and HPSEC were 1:4, 1:2, I:], 2: 1 and 3:l. For the IEF experiments, hGH and the monoclonal IgG were dissolved in 0.05 A4 potassium phosphate buffer, pH 7.6. The final concentration of IgG was 2.2 mg/ml, and that of hGH was 1.Omg/ml. Two solutions containing IgG and hGH were made by mixing appropriate volumes of the above protein solutions. The approximate molar ratios of 1gG:hGH in the above mixtures were calculated to be 1:1.5 and 1:3.1. RESULTS
The model antigen (hGH) and the monoclonal anti-hGH antibody (IgG) were readily available in highly purified preparations. They were well characterized in-house to establish purity, identity and binding properties. Results [14] showed that both preparations contained a single protein component of appropriate molecular mass by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) (hGH cu. 22 124 dalton, IgG cu. 150 000 dalton). No significant protein impurities were detected when stained with Coomassie Blue R-350. The apparent affinity constant for the reaction (&) was cu. 1.0 . IO9 l/mol by cold competition radioimmunoassay [ 143.This strong binding affinity of the antibody for hGH results in strong, stable antibody-antigen complexes that form very rapidly in solution. Mixtures containing various molar ratios of 1gG:hGH were analyzed by CZE, IEF and HPSEC to study the properties of the complexes formed. These techniques separate species based on differences in electrophoretic mobility, isoelectric point and size, respectively.
R. G. NIELSEN et al.
180
0
IOO 200
300 400
500 600
700 600
900
Time(s) B
I
Free IGG in mixture), ‘1,
II Ii
IGG:HGH=3:1 IGG:HGH
= 2:l
IGG:HGH
= 19
IQGG
0
150
300
450
El
600
750
900
Time (s)
Fig. I. Separation of IgG, hGH and IgG-(hGH), complexes by CZE. (A) Electropherograms of IgG, hGH and mixtures containing an excess of hGH. (B) Electropherograms of IgG, hGH and mixtures containing an excess of IgG. Experimental conditions are described in the text.
CZE experiments Fig. 1 shows seven electropherograms (El-E7) representing separate injections. Electropherograms El and E2 are for individual injections of IgG and hGH, respectively. They show that these species have different migration times and are well resolved under the experimental conditions employed. Note that TgG migrates much faster than hGH, due to the fact that the plfor hGH is 5.2 and that for the monoclonal IgG is in the range 7.2-8.0 (multiple bands in Fig. 2). Therefore, hGH molecules have much greater negative charge than IgG molecules in the CZE separation buffer (0.1 M tricine, pH 8). This results in faster migration of the IgG molecules towards the negatively polarized detector end of the capillary. The first peak in all electropherograms is due to a slightly higher tricine concentration in the sample buffer compared to that of the separation buffer. Electropherograms E3 and E4 show the results for 1:4 and 1:2 molar ratios of IgG:hGH, respectively (Fig. 1A). Based on migration-time comparison, the peak at CL~. 531 s is a new molecular entity, most probably corresponding to one or more IgG+hGH), complexes. The peak at ea. 576 s corresponds to free hGH present in excess in both mixtures. As the 1gG:hGH molar ratios decrease in the mixtures, one observes distinct changes in the peak profiles (Fig. lB, electropherograms E5, E6 and E7). In electropherogram E5, the free IgG peak was observed at CLI.495 s along with two distinct peaks for IgG-(hGH), complexes at cu. 513 and ea. 531 s, respectively.
CZE, IEF AND HPSEC OF ANTIBODY-ANTIGEN
COMPLEXES
181
Fig. 2. Separation of IgG, hGH and IgG
